36 Commits

Author	SHA1	Message	Date
Daniel Povey	1669e21c0c	Use decorrelation in conformer layers also	2022-06-09 11:31:52 +08:00
Daniel Povey	b9a476c7bb	Remove loss factor from decorr_loss_scale	2022-06-08 20:19:17 +08:00
Daniel Povey	8e56445c70	Try to resolve graph-freed problem	2022-06-08 20:07:35 +08:00
Daniel Povey	46ca1cd4c4	Add Decorrelate module that adds something to gradients in backward pass	2022-06-08 19:44:58 +08:00
Daniel Povey	9fb8645168	Implement JoinDropout	2022-06-08 16:17:42 +08:00
Daniel Povey	e7886d49a9	Bug fix	2022-06-08 11:05:29 +08:00
Daniel Povey	a83bde1372	Simplify implementation as current idea was not working to decorrelate	2022-06-08 10:24:41 +08:00
Daniel Povey	135be1e19c	Change dropout_rate from 0.2 to 0.1; fix logging statement; fix assignment to rand_scales, nonrand_scales to use [:]	2022-06-08 00:42:04 +08:00
Daniel Povey	a6050cb2de	Implement new, more principled but maybe slower version.	2022-06-07 23:38:38 +08:00
Daniel Povey	75c822c7e9	Pre and post-multiply by inv_sqrt_stddev,stddev	2022-06-07 20:32:18 +08:00
Daniel Povey	a270973b69	Add gaussian version of decorrelation	2022-06-07 18:55:48 +08:00
Daniel Povey	5d24489752	Have 2 scales on dropout	2022-06-07 18:31:42 +08:00
Daniel Povey	cd6b707e2b	Various bug fixes	2022-06-07 16:45:32 +08:00
Daniel Povey	40a0934b4e	Implement GaussProjDrop	2022-06-07 11:51:24 +08:00
Daniel Povey	4352a16f57	Fix bug that relates to modifying U in place	2022-06-06 17:43:15 +08:00
Daniel Povey	31848dcd11	Randomize the projections	2022-06-06 16:07:28 +08:00
Daniel Povey	6fdb356315	Bug fix RE GPU device	2022-06-06 15:40:20 +08:00
Daniel Povey	71e927411a	Implement FixedProjDrop	2022-06-06 15:38:59 +08:00
Daniel Povey	28df3ba43f	Fix bug re half precision	2022-06-05 23:26:59 +08:00
Daniel Povey	d76aedb790	Make it work for half	2022-06-05 23:25:51 +08:00
Daniel Povey	e535887abb	Bug fixes.	2022-06-05 23:24:02 +08:00
Daniel Povey	136ffb0597	Add ProjDrop for axis-independent dropout	2022-06-05 23:00:48 +08:00
Fangjun Kuang	f6ce135608	Various fixes to support torch script. (#371) ... * Various fixes to support torch script. * Add tests to ensure that the model is torch scriptable. * Update tests.	2022-05-16 21:46:59 +08:00
Guo Liyong	78418ac37c	fix comments	2022-04-13 13:09:24 +08:00
Mingshuang Luo	93c60a9d30	Code style check for librispeech pruned transducer stateless2 (#308)	2022-04-11 22:15:18 +08:00
Daniel Povey	2545237eb3	Changing initial_speed from 0.25 to 01	2022-04-05 18:00:54 +08:00
Daniel Povey	179d0605ea	Change initialization to 0.25	2022-04-04 23:34:39 +08:00
Daniel Povey	72f4a673b1	First draft of new approach to learning rates + init	2022-04-04 20:21:34 +08:00
Daniel Povey	37ab0bcfa5	Reduce speed of some components	2022-03-30 11:46:23 +08:00
Daniel Povey	11a04c50ae	Change 0.025,0.05 to 0.01 in initializations	2022-03-21 21:29:24 +08:00
Daniel Povey	8cff994cd7	Set also scale for embedding to 0.025.	2022-03-18 21:30:05 +08:00
Daniel Povey	188eada7ac	Change initial std from 0.05 to 0.025.	2022-03-18 21:28:34 +08:00
Daniel Povey	c9f1aeb7d1	Fix bug with import	2022-03-18 16:40:24 +08:00
Daniel Povey	ba3611cefd	Cosmetic changes to swish	2022-03-18 16:35:48 +08:00
Daniel Povey	6769087d70	Remove scale_speed, make swish deriv more efficient.	2022-03-18 16:31:25 +08:00
Daniel Povey	11bea4513e	Add remaining files in pruned_transducer_stateless2	2022-03-17 11:17:52 +08:00